On 3 Jul, 06:07, Rune Allnor <all...@tele.ntnu.no> wrote:

Bad typo needs correction:

> Of course, that doesn't prevent anyone from *modeling*
> the human speaker as a -s-o-u-n-d- source;

... model the human speaker as a *point* source ...

> that way all those
> difficult factors like directivity functions etc disappear from
> the equations. The risk is that there results are wrong
> since posibly important aspects of the real world are
> missing from the mathematical model.

Rune

On 3 Jul, 08:17, "Ron N." <rhnlo...@yahoo.com> wrote:
> On Jul 2, 9:07 pm, Rune Allnor <all...@tele.ntnu.no> wrote:
>
> > > > In source localization you very seldom have the oportunity to choose
> > > > the characteristics of the source. If you do, it almost always means
> > > > that you have access to the source and can find its position by
> > > > other means than acoustic localization techniques.
>
> > > Even with the increasing interest in small-area localization
> > > systems?  I think that in such systems one can "design"
> > > the sound source to a larger extent than what you said.
>
> > Interest in a system has nothing to do with feasibility of the same
> > system.
>
> > If you cand *design* a sound source, you have at your power to
> > include some tracking device like GPS or something like that.
>
> I doubt a GPS locator would work well for tracking a movable
> object in a cave, or even underwater, for instance.

..which is why I added "... or something like that." GPS doesn't work
underwater, but where I work we use tracking devices which are
surprisingly accurate in 1000 m deep waters.

Rune

On Jul 2, 9:07 pm, Rune Allnor <all...@tele.ntnu.no> wrote:
> > > In source localization you very seldom have the oportunity to choose
> > > the characteristics of the source. If you do, it almost always means
> > > that you have access to the source and can find its position by
> > > other means than acoustic localization techniques.
>
> > Even with the increasing interest in small-area localization
> > systems?  I think that in such systems one can "design"
> > the sound source to a larger extent than what you said.
>
> Interest in a system has nothing to do with feasibility of the same
> system.
>
> If you cand *design* a sound source, you have at your power to
> include some tracking device like GPS or something like that.

I doubt a GPS locator would work well for tracking a movable
object in a cave, or even underwater, for instance.

On 2 Jul, 22:53, julius <juli...@gmail.com> wrote:
> On Jul 2, 9:06 am, Rune Allnor <all...@tele.ntnu.no> wrote:
>
> > On 2 Jul, 11:59, "Sylvia" <sylvia.za...@gmail.com> wrote:
>
> > > In single source sound localization,which speakers are better to use in
> > > reverberant environment?Point source speakers or normal speakers.i would
> > > appreciate any list of standard speakers used in sound localization.
>
> > In source localization you very seldom have the oportunity to choose
> > the characteristics of the source. If you do, it almost always means
> > that you have access to the source and can find its position by
> > other means than acoustic localization techniques.
>
> Even with the increasing interest in small-area localization
> systems?  I think that in such systems one can "design"
> the sound source to a larger extent than what you said.

Interest in a system has nothing to do with feasibility of the same
system.

If you cand *design* a sound source, you have at your power to
include some tracking device like GPS or something like that.
If the source objects (the object which makes the sound) has
one of those onboard, it can report by radio its position as
measured by GPS, and sound localization is reduced to
merely an academic excercise. If the object is stationary,
one measures its position once and use tabulated values
ever after.

If, on the other hand, one is interested in "sources of
opportunity" -- human speakers in a conference room,
mammals or submarines at sea, birds outdoor -- there is
nothing one can do to "design" the source. The anatomy
of  ahuman being is and will remain what it is: The mouth
is the main sound source and the directivity is a major
factor. There is no way one can "design" a human to
alter those properties, hence one can not "design"
a human speaker into a point source.

That's a fact one just have to accept and deal with.

Of course, that doesn't prevent anyone from *modeling*
the human speaker as a sound source; that way all those
difficult factors like directivity functions etc disappear from
the equations. The risk is that there results are wrong
since posibly important aspects of the real world are
missing from the mathematical model.

Rune

On Jul 2, 9:06 am, Rune Allnor <all...@tele.ntnu.no> wrote:
> On 2 Jul, 11:59, "Sylvia" <sylvia.za...@gmail.com> wrote:
>
> > In single source sound localization,which speakers are better to use in
> > reverberant environment?Point source speakers or normal speakers.i would
> > appreciate any list of standard speakers used in sound localization.
>
> In source localization you very seldom have the oportunity to choose
> the characteristics of the source. If you do, it almost always means
> that you have access to the source and can find its position by
> other means than acoustic localization techniques.
>

Even with the increasing interest in small-area localization
systems?  I think that in such systems one can "design"
the sound source to a larger extent than what you said.

Now, interfering noise is a separate issue altogether, it's
not typical that one can control it.

Julius

On 2 Jul, 11:59, "Sylvia" <sylvia.za...@gmail.com> wrote:
> In single source sound localization,which speakers are better to use in
> reverberant environment?Point source speakers or normal speakers.i would
> appreciate any list of standard speakers used in sound localization.

In source localization you very seldom have the oportunity to choose
the characteristics of the source. If you do, it almost always means
that you have access to the source and can find its position by
other means than acoustic localization techniques.

So you basically have to cope with whatever is out there.

Having said that, point sources are very convenient mathematical
tools for modeling and analysis, as they are vanishingly small and
have uniform directivity. The point source model is often valid if
the source and reciever are far away form each other, i.e. when
the apparent size of the source when seen (or heard) from the
reciever is very small.

In practice, sources are not infinitely small and have non-uniform
directivity functions. These are characteristics one just have to
accept and try to handle as best one can, when dealing with
real-life measurements.

Rune